Hi everyone. I'm new here and don't really come from a strong math background which is why I found my way here.
Some background on what I'm doing: I play League of Legends with my friends and I coded an IRC bot that calls the Riot API to gather data on the match we are playing. The game is played with 2 teams 5 vs 5. When we enter game, the bot will report our chances of winning the game based on the win/loss of each champion. Currently I'm using a logistic regression model that I took from online calculator that does the same thing as I'm doing except that it's a webpage, not an IRC bot.
Like I said, my math background isn't extensive so I basically implemented something I really don't understand. All I know is that each champion has a number between -1 and 1 that indicates chance of winning.
I'm looking for a new model to calculate team win %. I store results in my database so my data is always increasing which I would like to impact the calculations. Also I'd like to factor in a new variable which is player rank. In League of Legends, players who choose to participate can play the came competitively. They are assigned a rank which increases or decreases based on their performance. I'd like to adjust the champion win percentage based on the player rank.
Example: If a champion has win rate of 52%, when the same champion was played by a high ranked player will have an effective win rate of 57%. I would like that to be factored into the team win percentage.
Here's an example of output of the current code for the IRC bot to help give context on what I'm doing:
beeetee is currently playing game mode: ARAM with a 75.1% of winning.
BLUE TEAM (24.9%): bagel saucee - unranked - Lissandra, FangedRabbit - unranked - Xerath, A Through Z - GOLD II - Xayah, ràt girl - SILVER I - Kaisa, Byakuun - PLATINUM IV - Leona
RED team (75.1%): A WHOLE GALLON - GOLD II - Jhin, Newgoodsdelivery - unranked - Veigar, CheeseCakePanda - BRONZE II - Graves, Beeetee - unranked - Chogath, GuardianOfMilady - unranked - Karthus
I have number of wins/losses for each champion in every game my friends and I play in.
Some background on what I'm doing: I play League of Legends with my friends and I coded an IRC bot that calls the Riot API to gather data on the match we are playing. The game is played with 2 teams 5 vs 5. When we enter game, the bot will report our chances of winning the game based on the win/loss of each champion. Currently I'm using a logistic regression model that I took from online calculator that does the same thing as I'm doing except that it's a webpage, not an IRC bot.
Like I said, my math background isn't extensive so I basically implemented something I really don't understand. All I know is that each champion has a number between -1 and 1 that indicates chance of winning.
I'm looking for a new model to calculate team win %. I store results in my database so my data is always increasing which I would like to impact the calculations. Also I'd like to factor in a new variable which is player rank. In League of Legends, players who choose to participate can play the came competitively. They are assigned a rank which increases or decreases based on their performance. I'd like to adjust the champion win percentage based on the player rank.
Example: If a champion has win rate of 52%, when the same champion was played by a high ranked player will have an effective win rate of 57%. I would like that to be factored into the team win percentage.
Here's an example of output of the current code for the IRC bot to help give context on what I'm doing:
beeetee is currently playing game mode: ARAM with a 75.1% of winning.
BLUE TEAM (24.9%): bagel saucee - unranked - Lissandra, FangedRabbit - unranked - Xerath, A Through Z - GOLD II - Xayah, ràt girl - SILVER I - Kaisa, Byakuun - PLATINUM IV - Leona
RED team (75.1%): A WHOLE GALLON - GOLD II - Jhin, Newgoodsdelivery - unranked - Veigar, CheeseCakePanda - BRONZE II - Graves, Beeetee - unranked - Chogath, GuardianOfMilady - unranked - Karthus
I have number of wins/losses for each champion in every game my friends and I play in.